sql
Pass
Audited by Gen Agent Trust Hub on Jul 16, 2026
Risk Level: SAFECOMMAND_EXECUTIONCREDENTIALS_UNSAFEPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides instructions and examples for executing shell commands via
docker compose exec mariadb mysql. This enables the agent to interact directly with the database server using CLI tools. - [CREDENTIALS_UNSAFE]: The agent is instructed to read database credentials from a local configuration file (
CLAUDE.local.md). Additionally, the recommended command pattern involves passing passwords via the-pcommand-line flag, which can expose secrets to other users on the system via the process list. - [PROMPT_INJECTION]: The skill processes data from a database (SQL query results), which constitutes an indirect prompt injection surface. Malicious data stored in the database could potentially influence the agent's behavior.
- Ingestion points: Output from the
mysqlCLI generated by commands inSKILL.md. - Boundary markers: None are specified; the agent displays results directly to the user.
- Capability inventory: The agent is capable of shell command execution and local file access as defined in
SKILL.md. - Sanitization: The skill does not implement validation or sanitization of the database output before presenting it to the agent's context.
Audit Metadata